• State TI Thinks Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 2 answers
  • Subscribers 63 subscribers
  • Views 126 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LM5033: Switching noise

John Sigve Risanger
Prodigy 20 points
Part Number: LM5033

Tool/software:

I'm using the LM5033 to modulate a DC voltage for transformer-based isolation, serving as part of a safety barrier mechanism.
However, we're observing switching noise that degrades EMI performance, and we're looking to mitigate it.

Currently, OUT1 and OUT2 are configured to switch at approximately 160 kHz, supporting a load of around 1 A. The system is powered with 24 V at VIN and 12 V at VCC.

We believe we’ve adhered to the recommended layout guidelines and have implemented several mitigation techniques: decoupling capacitors are in place, snubber networks have been added on the transistor gates connected to OUT1 and OUT2, and series gate resistors are also included.

We're considering whether adding a ferrite bead on the VIN line could improve EMI performance. Based on your experience, have you found this approach to be effective? If so, could you recommend a specific ferrite bead model? We're also open to any other improvement suggestions you might have.

  • 0
    TI__Guru 73495 points

    Hi,

    You need to measure the noise / spikes and find what their frequency and magnitudes then according to these to help select a ferrite bead.

    Layout is another area to look at to see any improvement can be made.

  • 0 in reply to Hong Huang
    Prodigy 20 points

    Of course.  I did write "switching noise".  The frequencies that makes trouble is the switcher frequency and its harmonics. 160kHz, 320kHz...
    My first and most important question was if anyone have seen any effect of adding a ferrite bead to the VIN line for such an issue, or is it maybe better to focus on the snubber network? Thanks.

  • 0 in reply to John Sigve Risanger
    TI__Guru 73495 points

    Hi,

    Ferrite bead is often used along with snubber but if snubber is ok to resolve the EMI issue without ferrite bead then it would be preferred.

    I think you need measure the noise frequency to see the highest magnitude noise frequency - it sounds you do not know but just logically thinking. You would need to know the frequency of highest magnitude noise to help select a ferrite bead.

  • 0 in reply to Hong Huang
    TI__Guru 73495 points

    Hi,

    Ferrite bead is often used along with snubber but if snubber is ok to resolve the EMI issue without ferrite bead then it would be preferred.

    I think you need measure the noise frequency to see the highest magnitude noise frequency - it sounds you do not know but just logically thinking. You would need to know the frequency of highest magnitude noise to help select a ferrite bead.

  • 0 in reply to Hong Huang
    Prodigy 20 points

    Thanks, of course I have measured :-)
    The #1 frequency is the 160 kHz.
    Thank you for trying to help me out, but please, I don't need any general answers.  This is the question from my initial post I need answer to: "We're considering whether adding a ferrite bead on the VIN power line could improve EMI performance.  Based on your experience, have you found this approach to be effective?"
    Actually I was hoping technical support from TI was answering out the question.

  • 0 in reply to John Sigve Risanger
    Prodigy 20 points

    Hong, My sincere apologies for overlooking the fact that you're part of the TI Tech Support team.

    It would be really helpful if future responses could more directly address the specific issue I’m facing. Thank you for your support and understanding!

    We have done investigations prior to this inquiry so we have pretty good control of the design utilizing the LM5033, but we are left with the mentioned switching noise.  Might be we still can optimize the layout or snubber network.  However, we wonder if possibly a filter on the VIN power line might have any impact.  

  • 0 in reply to John Sigve Risanger
    TI__Guru 73495 points

    Hi,

    Please provide your switching waveform so I can help further. I am not sure what is 160kHz frequency - which it looks it is a converter operation frequency which usually not able to use to ferrite bead to reduce its magnitude. Ferrite bead usually is used for much higher frequency noise like mega Hz on top of 160kHz.

    It is helpful you show the waveforms to help understand your issue to resolve - as it seems I cannot understand what you want to make attenuation as your description is not in line as I know in general. So the waveforms can help my understanding.

  • 0 in reply to Hong Huang
    Prodigy 20 points

    The LM5033 is a push-pull, voltage-mode PWM controller that provides alternating outputs (OUT1/OUT2) to drive external MOSFETs. These outputs are square wave signals set to operate at 160 kHz with a 50% duty cycle in my circuit.

    As with most switching regulators, switching noise is a known challenge during EMC testing. In this circuit, the 160 kHz switching frequency and its harmonics contribute to the EMI observed on the power line.

    However, after reviewing application notes and other documentation from TI regarding the LM5033, I haven’t seen ferrite bead on VIN being recommended. That’s why I initially brought this up. Let’s close this topic for now.  I’ll continue investigating it on my own. Thanks.